The story of Greek tragedy in the fifth century BC is an extraordinarily difficult one to tell. On the one side there are thirty-two well-known plays transmitted from antiquity through the medieval tradition, plays that have exerted a profound, even immeasurable, influence on Western culture, while on the other there are fragmentary scraps of evidence, often enough distorted by the preconceptions of later times, from which scholars try to reconstruct a whole history of an institution. How Dionysiac festivals were organised, what the earliest theatres, masks and costumes looked like, how the music sounded, what sort of performance-styles and dramatic conventions developed, how far the surviving plays are typical of the hundreds, or thousands, that must have been composed during the period, and what tragedy meant for the contemporary Athenian - and non-Athenian - audiences that watched it: these are the questions that need answers. What is lacking is systematic documentation, surviving from the fifth century itself, of this new and extremely successful artistic and civic phenomenon, and there is no prospect that anything of the kind will ever be recovered.

The best that modern research can hope for is new fragments of evidence - a vase-painting or an inscription, a papyrus text of part of a lost play or of a scholar's introduction (hypothesis) - which will fill some of the gaps in the story. The most striking example was the publication in 1952 of a small papyrus scrap of a hypothesis which proved that Aeschylus' Suppliant Women was not the earliest surviving Greek tragedy but belonged to the 460s, and therefore to a late stage in the poet's career. This play, with its chorus of the daughters of Danaus (the myth said there were fifty of them), had previously been taken as a sample of the tragedy of the 490s and was thought to have a chorus of fifty like the dithyramb; it was read as a 'primitive' piece more akin to choral lyric poetry than to the true dialectic of drama.